How to Diagnose Underperforming Client Campaigns in 10 Minutes Every Morning
Alessandro Conti
Senior Performance Marketer
The hardest moment in agency media buying is hearing a client say "I noticed my CPA doubled yesterday — what happened?" before you did. That conversation is recoverable, but it should not happen. This guide walks through how to diagnose underperforming client campaigns in 10 minutes every morning — a structured ritual powered by Wevion anomaly alerts that puts the buyer in front of the problem before the client is.
Quick answer: Wire five threshold alerts per client account — CPA ceiling, ROAS floor, CPM spike, CTR collapse, budget pacing. Each morning, open the overnight alert list, filter by severity, confirm the signal, identify the root cause, and write a one-line note. Ten minutes covers a 20-plus account roster, and the buyer reaches every client call diagnosis-ready.
The morning audit ritual is not about working harder on the dashboard sweep. It is about replacing the sweep entirely — so the buyer's attention flows to the accounts that signaled a problem, and every other account stays quiet.
Why the dashboard sweep fails as a diagnostic tool
The dashboard sweep is the default because it feels thorough. Open each client account, scroll the metrics, look for anything that seems off. But "seems off" is not a diagnostic standard — it is pattern recognition under time pressure across a dozen platforms that all display data differently.
The sweep fails in two specific ways. First, it covers accounts sequentially, which means the buyer is comparing yesterday's numbers in their head while looking at today's — a cognitively expensive, error-prone translation that does not surface the delta cleanly. Second, it is a snapshot: the buyer checks at 9am and has no signal until they check again. A CPA that doubled at 11am runs unnoticed until the afternoon sweep or, worse, until a client calls.
The dashboard sweep is not monitoring — it is a scheduled guess. The buyer looks at every account and hopes something catches their eye, instead of having a system surface the accounts that crossed a threshold. Anomaly alerts replace guesswork with a signal: the account raises its hand, the buyer investigates, and everything else stays off the desk.
AgencyAnalytics' 2024 agency benchmarks identified reporting and account-monitoring as the largest recurring time sink for performance agencies, with multi-platform accounts amplifying the problem. The fix is not a better sweep; it is changing the model from pull to push — the platforms report to the buyer, instead of the buyer checking the platforms.
The five alerts that drive the morning audit
The morning ritual starts the night before, when the threshold rules are already doing their work. The five signals that surface genuinely underperforming campaigns — rather than normal variance — are:
CPA ceiling: cost per acquisition above the client-specific floor of profitability. This is the client's core business metric, and a breach means the campaign is burning spend without the expected return. Wire this per account, not as a universal threshold — a €15 CPA is a win for one client and a disaster for another.
ROAS floor: return on ad spend dropping below a minimum viable number for the client's margin structure. ROAS collapse is often a creative or audience signal before it becomes a budget signal — catching it early means the fix is a creative swap, not a campaign rebuild.
CPM spike: cost per thousand impressions rising sharply against a baseline. A CPM spike usually reflects an auction problem — increased competition, an audience segment becoming expensive, a placement change — and is the earliest leading indicator of margin erosion before CPA or ROAS moves.
CTR collapse: click-through rate falling against the account's historical norm. Creative fatigue shows up in CTR before it reaches CPA. A CTR collapse that is not caught early becomes a CPA problem by the end of the week.
Budget pacing: spend running materially ahead of or behind the expected daily curve. Ahead means a budget cap risk by midday; behind means the campaign is under-delivering against the client's flight plan and needs an investigation into delivery constraints.
Each alert represents a different failure mode with a different fix. A CPA breach points toward creative, audience, or bid. A CPM spike points toward the auction. A CTR collapse points toward creative fatigue. A pacing miss points toward delivery. Wiring all five gives the buyer a structured diagnostic path — not an open-ended investigation from a bad-looking number.
The 10-minute morning ritual: step by step
With the five alerts wired, the morning audit becomes a structured, time-boxed routine rather than an open-ended sweep.
Minute 0–2: Open the alert list, filter by severity. The overnight alert queue shows every threshold breach across all client accounts since the last check. Sort by severity — a CPA breach at two times the ceiling outranks a minor pacing variance. Identify which accounts need investigation and which are informational. On a typical morning, this narrows 20-plus accounts to three or four that need attention.
Minute 2–6: Confirm the signal per flagged account. For each flagged campaign, open the account view and verify the alert reflects a real drift rather than a data artifact. Check the metric trend over the last 24 to 48 hours: is the CPA spike a sustained move or a single-hour anomaly? Is the ROAS floor breach growing or recovering? Confirming the signal takes one to two minutes per account and prevents the buyer from surfacing a false alarm to the client.
Minute 6–9: Identify the root cause. With the signal confirmed, the structure of the alert points toward the most likely cause. A CPA breach with a CPM spike alongside it suggests an auction shift. A CPA breach with CTR flat suggests a conversion-side problem — landing page, offer, audience match. A ROAS collapse with stable CTR suggests the purchase value or attribution is the issue, not delivery. Cross-referencing two or three metrics in the account view gives the buyer a working hypothesis in under two minutes.
Minute 9–10: Prepare the one-line status note. For each flagged account, write one line: the metric breached, the current value, the hypothesis for root cause, and a proposed next action. This is the brief the buyer carries into the client call or the morning team standup. The client hears "CPA ran to €38 overnight against your €25 ceiling — looks like a CPM spike on the broad audience; we're testing a narrowed lookalike today" instead of "we're looking into it."
The 10-minute ritual is not about speed for its own sake. It is about arriving at the client conversation with a diagnosis rather than a question. The buyer who says "here is what happened and here is what we are testing" is in a far stronger position than the buyer still investigating when the client calls.
Routing alerts so every flag reaches the right buyer
On a multi-buyer team, the routing layer is what makes the ritual scale. An alert for a client the buyer does not manage is noise; an alert for their account is signal. Spray every account's overnight queue at the whole team and the channel is muted within a week.
Route each client account's alerts to the buyer responsible for that account. For spend-level and budget alerts that affect the agency's billing relationship, the lead or account manager may also receive them. The result is a Telegram channel — or whichever delivery channel the agency uses — that is quiet most of the time and actionable every time it fires.
The mechanics of routing across a multi-account team are covered in detail in agency ad alerts across many client accounts, and the alert delivery channel comparison is at ad alert delivery: Telegram, email, in-app compared.
What happens when the buyer finds the problem first
The shift in the client relationship is the point of the whole ritual. When the buyer calls the client at 9:30 with "we caught a CPA drift overnight and we have a test running this morning," the client's experience of the agency is competence and attention. When the client calls the buyer at 10 with "did you see my CPA?" the agency is reactive and the trust is eroded, even if the fix is identical.
According to HubSpot's 2024 Agency Trends Report, client communication and proactive reporting rank among the top three drivers of agency retention. The morning audit ritual is an operational answer to a retention problem: the buyer is structurally positioned to surface issues before the client does, which makes the client feel watched rather than watching.
Catching a client's underperforming campaign before they do is not just an operations win — it is the evidence that the agency's attention is always on. The 10-minute morning ritual, powered by overnight anomaly alerts, turns a monitoring process into a client-trust asset: the buyer shows they were already on it.
Build the ritual into the team's daily flow
The ritual scales across team size and account count with the right setup: alerts wired per account, routing scoped per buyer, a shared morning window where each buyer works their queue, and a team standup where flagged accounts get a 30-second status.
For agencies ready to wire the full morning monitoring stack — anomaly alerts, budget pacing, creative signal tracking, all in one view — the agency-tools cluster hub holds the full playbook. Start with the alert wiring guide, and add budget pacing checks as the second layer of the morning ritual.
Agencies running multi-client monitoring typically operate on Pro €499 or Plus €1,499/month (€1,199 annual), with Enterprise custom for the largest shops. The 14-day trial alongside the permanent free tier is enough to wire the five alerts, run the morning ritual for a week, and see whether you are the one calling the client or the client is calling you.
Frequently Asked Questions
The Ad Signal
Weekly insights for media buyers who refuse to guess. One email. Only signal.
Related Articles
How One Agency Stopped Babysitting 14 Client Accounts With Phone Alerts
A four-person agency was losing the first two hours of every day to a dashboard sweep across 14 client accounts on five platforms. This is the story of how routed Telegram alerts changed that — which alerts they wired, how they split them across the team, and why the line between notify and act mattered as much as the time they saved.
9 Ad Alerts That Actually Matter (And the Ones to Mute)
An alert channel is only as good as the alerts you put in it. Wire the wrong ones and you train yourself to ignore the right ones. This is the field-tested list of nine cross-platform ad alerts worth a phone notification — and the common ones to demote to a digest before they drown the signal.
Agency Daily Budget Pacing Across Multiple Clients: Replace Manual Checks With Automated Rules
Checking daily budget pacing across 20-plus client accounts by hand is the kind of work that looks thorough but leaves gaps. This guide shows how an agency replaces those manual sweeps with automated pacing rules and threshold alerts — so overspend and underspend surface in minutes, not at the end of the flight when it is too late to fix.